Healthcare data saves lives. It is a proven fact that digitising a patient's health record will improve the overall efficiency, reach and quality of healthcare delivery manifolds. However, manpower shortage, less insurance penetration and the skewed doctor to patients ratio are major roadblocks to digitisation in India. During pandemic outbreaks, information based on provisional diagnoses as reported by clinicians is pivotal to direct the outbreak control efforts in the right direction. That said, of all the data types in medical informatics (imaging, etc.), doctors’ prescriptions are still handwritten which significantly slows the transmission of data up the chain of command and the decision making cycle. 

While doing his research in precision medicine and oncology informatics, Mumbai based Randeep Singh understood the struggle of doctors on data entry and recording healthcare data. Singh needed a lot of data to produce any meaningful findings or to begin therapies. However, digitized and structured data was nowhere to be found. Everything was on paper. To sort out this issue and to speed up health digitisation he found a startup, Doxper in 2015 and came up with a digital pen and encoded paper solution

AI-powered Pen

Powered by cloud computing and machine learning, Doxper’s digital pen and encoded paper is a clinical documentation system. Pens and papers in a hospital are replaced with Doxper pens and encoded papers and doctors, and other medicare professionals should write the patient’s data using the Doxper pen on the encoded papers. 

“The AI-powered digital pen and encoded paper solution that digitises the doctor's handwritten strokes in real-time through BlueTooth. We process the data and connect it to a larger ecosystem, unlocking benefits for patients, administrators and policymakers, “ says Parag Agarwal, Head Partnerships of the company. “ The AI in our system is a proprietary Machine Learning algorithm that can transcribe handwritten strokes with a very high level of accuracy, and the system continuously learns with more context and data. It is the first and only clinical data capture system that comprises a digital pen and encoded paper,” he says. 

Existing solutions such as typing, tapping or speaking increase consult times, cognitive load and physician burnout, while also reducing attention for the patient. One of the major benefits of the solution is that it does not require any behaviour change from the doctor/ nurse/ assistant and does not require any typing. Hospitals also do not have to change their workflows, while benefiting from very high data compliance and tracking and plugging of revenue leakage. 

The Doxper workflow is patent pending with proprietary handwriting recognition & clinical NLP algorithms.  

Tracking COVID-19

The solution can be quickly deployed to all of the COVID-19 screening and testing centres, to ensure that real-time, accurate data is digitised and flows to the right stakeholders for updates and analysis says the company. Additionally, in the case of false negatives or new cases discovered, a digitised paper trail will allow for rapid contact tracing and investigation. “In public health, we are detecting and deterring fraud, tracking performance and ensuring overall accountability and traceability of data,” says Agarwal.

Currently, the company is serving the solution to 2000 clinics and 25 hospitals including Nanavati, Hinduja hospitals in Mumbai, KIMS Hyderabad, Cloudnine and Wockhardt Hospitals chains. 15 more hospitals are in the pipeline to implement the solution. “In public health, we currently work on a project basis. We are processing 600,000 prescriptions every month,” says Agarwal. 

Since this is a cloud-based product and offers subscription-based pricing models, the solution is very affordable for hospitals, claims the company. It is also based on a modular architecture, cloud-hosted and without any silos. “Our goal is to not only digitise but to connect the country’s health infrastructure across public and private thus making data transmission secure and seamless. This will also provide more data for administrators, policymakers and researchers in real-time,” stated Agarwal.

Doxper is also focusing on natural language processing and working more intelligently with both structured and unstructured data.